1

A MIGS OCCASIONAL PAPER

Kenya: Violence, hate speech and vernacular radio by Keith Somerville,* Lecturer in Journalism, BA (Honours) Journalism Convenor, Brunel University, London (Keith.Somerville@brunel.ac.uk) (Revised manuscript submitted 2 March 2010)

In late December 2007, the delays in the ballot count and the controversy over the

announcement that Mwai Kibaki had been re-elected as President of Kenya sparked

an outbreak of violence which mounted in intensity and brutality and continued for

months. Between 1,200 and 1,500 people died and over 660,000 were displaced

from their homes and localities.

What started as demonstrations by the main opposition movement and its supporters

became widespread and in places violent, with the police and security forces

reacting with lethal force. Protest and violence, clearly sparked by the election,

rapidly took on the appearance of an ethnic struggle between Luo and Kalenjin

supporters of Raila Odinga and the Orange Democratic Movement (ODM) and

mainly Kikuyu supporters or perceived supporters of Kibaki‘s Party of National Unity

(PNU). As the violence escalated the planned and organised nature of the attacks

on particular communities became clear – this was not a spontaneous outburst of

anger, but carefully orchestrated violence with clear political and economic

objectives.

Within a few weeks, Kenya‘s post-election violence had spawned comparisons with

the 1994 genocide in Rwanda. Kenyan politicians on both sides of the electoral

divide exchanged accusations of genocide and ethnic cleansing, with little concern

2

for the facts, their own roles in the events, the consequences for Kenya or for the

image of their country or continent. The genocide /ethnic cleansing theme was

rapidly taken up by many Western and Kenyan journalists – there were worrying

comparisons being made by Kenyans and foreign observers alike with the 1994

genocide in Rwanda. Part of this comparison involved warnings that politicians and

community leaders were engaged in inciting ethnic and community hatred with a

view to encouraging violence and that hate radio was operating in Kenya and was at

least partly to blame for inter-communal violence and the inflaming of suspicion, fear

and hatred between ethnic groups. Human rights organisations, media monitors,

politicians and commentators all pointed to the role of the media, and vernacular

radio stations in particular, in inciting violence, contempt and ethnic hatred (see for

example IRIN 27January 2008).

The major accusation against a number of radio stations broadcasting to the

Kalenjin, Luo and Kikuyu communities in their own languages was that they were

deliberately and knowingly increasing ethnic suspicion, directly advocating violence

against ―others‖ and disseminating messages of hatred and incitement. The IPS

News Service immediately drew a comparison with Rwanda and the well

documented role of hate radio during the 1994 genocide:

The media was partly blamed for the Rwandan genocide 14 years ago which left nearly one million people dead in 100 days. "Kill the Inkotanyi [cockroaches]!" a local radio station urged its listeners at the time. "30 Days in Words and Pictures: Media Response in Kenya During the Election Crisis" - a workshop organised here last week by California-based media advocacy group Internews - enabled media professionals to conduct a "self-audit" of the role local media played in the post-election violence. The audit revealed that media - especially vernacular radio stations - might be partly to blame for the on-going violence sparked off by the announcement of Mwai Kibaki as winner of the Dec. 27 elections.(IPS, 2 February 2008)

3

Some experienced media observers, who had been monitoring Kenyan radio during

the elections and the violence, were more cautious but still drew parallels with

Rwanda and suggested that Kenya could move towards a situation in which radio

stations acted like the infamous Radio-Television Libre de Mille Collines (RTLM) in

Rwanda. Caesar Handa, the director of Strategic Public Relations Research, which

produced media monitoring reports for the UN Development Programme (UNDP),

said that, ―we did not reach the Radio Mille Collines level, but we were not very far

from it‖ (BBC World Service Trust, April 2008, p. 5).A chilling image. But how

accurate is this portrayal of the violence in Kenya and of the role of radio in

spreading hate and inciting violence? In looking at these questions this paper will:

examine the political/social environment that gave rise to the violence; locate it in the

historical context of political violence in Kenya in recent decades and in the language

of political discourse in Kenyan politics; examine the behaviour of the Kenyan media

and particularly vernacular radio during the elections and the ensuing violence;

compare Kenyan radio behaviour with RTLM in Rwanda in terms of dissemination of

fear, hatred and incitement to murder; and analyse the use of terms such a

genocide, ethnic cleansing, ethnic hatred and tribal animosity – which were widely

used in describing the violence by politicians, commentators and the media both in

Kenya and internationally.

Due to the absence of a large body of transcripts of vernacular radio from the period,

the analysis will be primarily qualitative rather than quantitative and for its

methodology will draw on elements of critical discourse analysis as applied to media

content. The author has interviewed Kenyan journalists, media monitors and

government press officials on the behaviour of vernacular radio and the media in

general during the period.

4

The political and media discourses that were part of the Kenyan political process and

which were accused by many of playing a major role in the violence will be analysed

using the criteria of representation (how social actors, events and institutions are

represented within the discourse), framing (how actors and events are

contextualized with a discourse, what level of prominence are they given) and

assumed meanings (Fairclough, 2003, pp.26-8, 53, 58). The analysis will work from

the basis that ―every single instance of language use reproduces or transforms

society and culture, including power relations...discourses are historical and can only

be understood in relation to their context‖ (Richardson, 2007, pp. 26-7).

Of particular relevance will be the extent to which radio stations and the messages

they broadcast had the effect of setting an agenda for their listeners over time and

especially at periods of heightened tensions and the way in which framed their

discourse and represented political/social actors. As Kellow and Steeves wrote in

their study of the role of radio in the Rwandan genocide, framing is about selection

and salience of content and in times of conflict or potential conflict might include

depiction of risk or danger to the audience from others, dramatization of the conflict

and inflation of the power or strength of opponents. Events and perceptions are

framed and agendas are identified and in this way for those engaging in the

broadcasting of hate messages; ―a media campaign is a conscious, structured

attempt to use media to influence awareness, attitudes or behaviour‖ (Kellow and

Steeves, 1998, p.111).

5

The political context: shifting alliances and violence as an

inescapable part of election campaigns:

The 2007 election was a contest between two main political rivals, the incumbent

President, Mwai Kibaki, and his Party of National Unity (PNU), and the opposition

coalition, the Orange Democratic Movement (ODM) led by Raila Odinga. Kalonzo

Musyoka of the smaller ODM-Kenya also fought the presidential election and a total

of 20 parties put up candidates for parliament. The bitterness of the campaign had

its roots in Kenya‘s developing political culture and, on a personal level, in events

following the previous election.

During the 2002 election, which saw Daniel arap Moi‘s favoured candidate Uhuru

Kenyatta defeated, Kibaki and Odinga had led the National Rainbow Coalition

(NRC), made up of politicians from a variety of parties, politicians and communities

opposed to Moi. A pre-election deal between them fell apart as the conservative

Kibaki used the presidency to entrench his own power and to favour his closest allies

at the expense of his alliance partners. By the time of the 2005 constitutional

referendum, Odinga had become a fierce critic of Kibaki. The referendum itself was

the result of a tortuous process of constitutional reform. Kibaki and Odinga had

agreed to institute this reform – long called for by many politicians in Kenya – when

they took power. Odinga believed that Kibaki had agreed to a diminution of

presidential powers and the creation of the post of prime minister with extensive

executive powers – a post which he expected to hold.

Reform proved hard to agree and it was two tortuous years before a draft was

agreed at what was called the Boma conference – the draft was broadly based on a

reduction in presidential powers and the creation of the post of prime minister.

6

Kibaki was not happy with this version and changed the draft prior to the November

referendum. This resulted in a fiercely fought and violent referendum campaign (in

which hundreds died and tens of thousands were displaced by violence). Odinga

forged an alliance of groups from Luo, Luahya and Kalenjin communities and

political leaders, which was in direct opposition to Kibaki and his allies. Odinga was

successful and Kibaki‘s draft was voted down with the ―No‖ vote registering 57

percent of the vote. Kibaki dissolved his cabinet, appointing a new one drawn from

the ―Yes‖ camp, irrevocably splitting the alliance which had defeated Moi. The

violence of the referendum campaign in the Rift Valley, Kisumu and Mombasa, and

the bitterness of the campaigning was a foretaste of what was to come in 2007.

The Kriegler report (commissioned after the 2007-8 violence by the Kenyan

government and by the international mediators who negotiated a political

compromise between Kibaki and Odinga, and chaired by the South African judge,

Justice Johann Kriegler) reported that the passions that developed during the

referendum campaign had maintained political discourse at a high pitch leading up to

the 2007 election campaign, which was characterised by ―robust language

occasionally lapsing into ethnic hate speech and deteriorating into violence‖

(Kriegler, 2008, p.1).

In addition to the controversy and bitterness over the referendum, Kibaki‘s

incumbency saw none of the redistribution of land, wealth or solving of historical land

and other grievances that Odinga‘s supporters had hoped for when they joined

together with him to form the National Rainbow Coalition. The president was seen

by many Kenyans as ruling for the benefit of a small group of politicians and

7

businessmen, who became known as the Mount Kenya Mafia. This group was

perceived by other communities - notably the Luo and Kalenjin - as working solely for

the benefit of a small Kikuyu/Kiamba elite closely allied to Kibaki and benefitting

through access to wealth, government contracts and political patronage (Anderson,

Royal African Society website, 2008).

As the 2007 elections approached, Odinga used the ―No‖ alliance of the

referendum to put together a coalition of politicians who supported a populist agenda

based on poverty reduction, the settlement of land and economic grievances dating

back to colonial land seizures, a revisiting of constitutional issues (including a

commitment to revisit federalism and devolution of power to provinces and even

lower levels – something which particularly appealed to Kalenjin and Masaai political

leaders) and a more equitable and redistributive economic strategy. For once, there

were not only personality issues between ―big men‖ to the fore in the election but

serious political and economic issues. There was also a major realignment of politics

and allegiances since 2002, with the victorious alliance splitting and becoming bitter

enemies, but also with many members of the former Moi government backing Kibaki

(Cheeseman, 2008, p.167).

The president‘s economic policies were seen by many Kenyans outside the political

and business elite as benefitting the rich and further impoverishing and marginalizing

the poor (IHT, 10 January 2008). There was even strong criticism within the Kikuyu

community of the corruption surrounding the Mount Kenya Mafia. During Kibaki‘s

period in office this had manifested itself in the growing power within poor Kikuyu

communities of a criminal gang/sect known as Mungiki. This group ran protection

8

rackets and carried out violent extortion in poor urban areas, recruiting from among

poor, badly educated and unemployed Kikuyu youths and using oathing ceremonies

reminiscent of the Kikuyu-based Mau Mau anti-colonial movement of the 1950s. It

came into conflict with the government and in the years running up to the 2007

election there were outbreaks of violence in slum areas and the killing of police and

sect members in clashes between the security forces and Mungiki. Nic Cheeseman

believes that just months before the elections up to 500 suspected Mungiki were

killed by government security forces in the slums of Nairobi – part of an estimated

600 persons who died between the start of the election campaigning in October 2008

and the vote at the end of December (Cheeseman, 2008, p.170). Despite the violent

struggle with the government prior to the elections, the Mungiki (along with similar

criminal gangs or youth gangs in other communities) became available as the paid

thugs for Kikuyu politicians during the elections.

The relationship between politicians and criminals or loosely-organized gangs of

poor and unemployed youths had been a growing facet of Kenyan politics since the

breakdown of the one-party system and the dilution of the monopoly of state power

enjoyed by the ruling party from independence up to the early 1990s – there had

been a growing ―informalisation‖ of political violence. There were no permanent

alliances involved here, as Branch and Cheeseman point out, but shifting

relationships based on money, mutual benefit of gangs and political barons and

common community/ethnic identity - the latter not specifically a result of ethnic

animosities but because gangs were recruited locally in the political

strongholds/heartlands of powerbrokers and politicians (Branch and Cheeseman,

2009, p.15). Former Moi loyalists like Nicholas Biwott and William Ntimama had

started developing their own army of bodyguards and thugs prior to the end of single

9

party rule in 1991 – many other politicians followed suit or had shifting relations with

gangs available for hire (Mueller, 2008, p. 189).

In terms of their support bases, Odinga, had Luo support and the support of younger,

poorer Kenyans, particularly in Nairobi slums like Kibera. His alliance forging was

assisted by a struggle for power within the political elite in the Rift Valley, the results

of which enabled him to gain the backing of the most powerful political factions in the

Rift Valley from among Kalenjin and Masaai communities. It was not a foregone

conclusion that the Kalenjin and Masaai would opt to support Odinga against Kibaki.

As Lynch has described (Lynch, 2008, pp. 542-3), the Kalenjin community has

usually voted en masse for the same party in a particular election – though not for

the same party in successive elections. The Kalenjin supported Moi‘s Kenya African

Democratic Union (KADU) in 1963 and the Kenyan African National Union (KANU),

by then led by Moi, in 1992, 1997 and 2002 (though in 2002 Moi had been barred by

the constitution from running for a third term and had supported Uhuru Kenyatta as

the KANU candidate, but was still the pre-eminent Kalenjin political leader). In the

latter election, Kalenjin leaders supported Moi against both Kibaki and Odinga.

The Rift Valley political changes in 2007 were the result of a number of factors,

including generational struggle, within the Kalenjin between Moi (who in 2007 was

supporting Kibaki, along with other Moi era political notables such as the Kalenjin

politician Nicholas Biwott and the Kikuyu politician and son of Kenya‘s first president,

Uhuru Kenyatta) and a younger politician, William Ruto. Ruto was a former KANU

10

MP and minister under Moi. Ruto won the struggle and threw Kalenjin support

behind Odinga and the ODM.

Ruto and his supporters wanted a future Odinga-led government to settle land issues

in the Rift Valley and move towards devolution of power away from the central

government in Nairobi – a process known as majimboism and long vilified in Kenyan

politics by the governing elite. In the 1960s and 1970s, President Kenyatta, his

supporters and those enjoying the fruits of government power depicted majimboism

as being ethnically/tribally-based and a threat to national unity. Kenyatta had

successfully co-opted the pro-federalist, Kalenjin leader, Daniel arap Moi, and his

supporters into government in 1964, thereby undercutting the federalists in the

1960s and rendering them marginal to mainstream politics.

The issue was largely dormant in mainstream politics during most of the Moi period,

as previously pro-federal politicians from the predominantly Kalenjin Rift Valley and

from Masaai areas (and to a lesser extent in other provinces away from Nairobi and

Kenyatta‘s Central province heartland), now had their hands on the levers of power

and sources of patronage. It only became a major political issue again when

demands for multiparty rule from Moi‘s opponents put him under political pressure.

Moi and his political lieutenants in KANU revived ideas of federalism. Majimboism

and the devolution of power to regional/provincial bodies, notably in the Rift Valley,

were seen by Kalenjin leaders as a possible way of preserving their power bases

and sources of economic wealth and political patronage in the face of threats to their

monopoly control of national politics represented by multiparty rule (Klopp, p.484).

11

Devolution became a central plank of the ODM campaign, though with Odinga trying

to avoid being tarred with the brush of a simplistic majimboism, with its connotations

for some Kenyans of ethnic separatism and ethnic cleansing. Moi‘s Kalenjin

supporters had been widely accused of carrying out ethnic cleansing in the Rift

Valley under the guise of majimboism in the 1990s, a period which saw over 2,000

deaths and 500,000 displaced in land conflicts in the Rift (Mueller, 2008, p. 191).

The political aspects of this approach to competitive politics were bound up with

grievances over land ownership and occupation dating back to land seizures by

white settlers during the colonial period and subsequent politically motivated

resettlement schemes after independence. Resettlement schemes launched by

Kenyatta were perceived by the Kalenjin to have benefited Kikuyu and Kisii migrants

into the Rift Valley rather than the Kalenjin or Masaai, who claimed original

ownership of the land (HRW, 2008, p.5). In the 1990s and 2002, the question of

land and the grievances of the Masaai and Kalenjin became key rallying points for

politicians allied with Moi and against what were seen as Kikuyu-led or influenced

parties/coalitions established as one-party rule was abolished. These issues were

exploited repeatedly during the 1992, 1997 and 2002 elections, with varying degrees

of success, but always with elements of whipped up anti-outsider feeling and the

threat or actual use of violence (see, for example, Klopp, pp.485-91and HRW, 2008).

This pattern was repeated in the 2005 referendum campaign. To cite John

Lonsdale, ―Violence arises from the fresh competition for the territory previously

allocated to African settlement, and for ethnic advantage, on the departure of white

settlers in the 1960s‖ (Lonsdale, 2008, p.308).

12

During the 2007 election campaign, beginning in November, violence was frequent

and serious and was accompanied by the effective barring of politicians and party

supporters from one alliance from areas that were the strongholds of opposing

parties. Election monitors, who were highly critical of the fairness of the elections

and the vote count, pointed to the inability, for example, of the ODM or ODM-Kenya

to campaign in pro-PNU areas of Central Province. Similarly it was impossible for

the PNU to campaign safely or effectively in areas of the Rift Valley and Nyanza

province (EUEO, 2008, p. 8). Politicians or activists campaigning in opposition

territory were liable to be physically attacked and have their rallies or meetings

broken up by opposition supporters or paid thugs.

Kenya has experienced serious outbreaks of politically-motivated violence at each of

the elections from 1992 onwards (1992 marking the resumption of multiparty

elections after a gap of 29 years), during the attempted coup against President

Daniel arap Moi by air force officers in 1982 and during the 2005 referendum

campaign. Violence has also punctuated disputes over land ownership, political

patronage and historical grievances derived from the upheavals of the colonial

period (for more detailed examinations of these issues see Klopp; Lynch, 2008;

Berman, 1990; Human Rights Watch 2008). But the ferocity, rapid escalation and

scale of the violence after the 2007 election took many Kenyans and international

observers by surprise, shattering myths of Kenya as an essentially politically stable

country (Guardian, 31 December 2007, Financial Times, 29 December 2007 and

Independent, 6 January 2008 all have variants on the ―haven of stability descends

into violent chaos‖ approach to reporting the violence).

13

Politicians intent on maintaining or extending their power and privileges or struggling

to challenge the powers of local or national rivals used a variety of instruments to

strengthen their own power bases and utilised violence and hate propaganda against

opponents and their opponents‘ real, perceived or potential supporters (Human

Rights Watch, 2008). They created an atmosphere in which they could mobilise

supporters to carry out violent attacks against opponents and in which their

supporters would, on occasion, react spontaneously to events, having been incited

to expect threats of fraud from their opponents.

In 2007, there was growing violence in the latter stages of the campaign and then a

substantial escalation following the announcement of the results and accusations by

losing candidate, Raila Odinga of the Orange Democratic Movement, that the count

in the presidential election had been fraudulent,—accusations backed up by reports

from the international observers, notably the European Union monitors (EUEO,

2008). The violence followed directly from the election and the disputed result,—

there had been no widespread violence in the areas affected prior to the start of the

election campaign.

The initial protests by ODM supporters in the Kibera slum near Nairobi, the Rift

Valley and Nyanza province drew a swift and brutal response from the police, with

fatalities resulting from police use of extreme force and live ammunition to combat

demonstrations (HRW, 2008, pp. 27-30). But the unrest generated by the ODM, the

police response and then the reaction by PNU supporters appeared to take on a life

of its own and rapidly came to resemble ethnic warfare – notably between

Luo/Kalenjin supporters on the one hand and Kikuyu/Kisii supporters of the PNU on

the other.

14

The growing violence and mounting deaths were widely interpreted by politicians in

Kenya (for their own mercenary purposes) and by the Kenyan and international

media as ethnic or ―tribal‖ violence that was a spontaneous reaction to the election

result that had its roots in historical and primordial tribal hatreds. Powerful politicians

who were using political violence as an instrument – whether ODM leaders using it to

force a recount or rerun of the election, or PNU leaders using violence to protect

their less than fair election victory – wanted the violence to be seen as spontaneous,

ethnic or ―tribal‖ violence to mask their own organising roles (Anderson, Prospect,

2008) and to use the ―tribal‖ aspect as a propaganda stick with which to beat their

opponents. Close examination of the violence shows that however spontaneous

some initial demonstrations were, the escalation and proliferation of violence was to

a great extent orchestrated by politicians and showed significant planning (HRW,

2008). For journalists, the ethnic or tribal explanation was a convenient, lazy and

value-laden shorthand which absolved them of the need to go into complex

explanations of the intricacies of Kenyan politics and generally fitted in with their

assumptions about the ethnic/tribal African politics (Somerville, 2009). Certainly, in

some cases local communities or party supporters reacted spontaneously to

suspected threats or to the heavy-handed suppression of demonstrations by the

police.

There are major problems with the tribal hatred approach. The Kalenjin shifted their

political allegiances, at times supporting Moi and his allies, at others Kibaki and

Odinga. And, prior to colonisation there was no such group as the Kalenjin per se.

15

The ―tribe‖ brings together ―a number of sub-groups administered as separate tribes

during the colonial period‖ (Lynch, 2008, p. 542). The term Kalenjin, first used in the

late 1950s, was a creation of the colonial period though not directly of the colonising

power, when it suited nascent political leaders from within the groups (including

Nandi, Kipsigis and Tugen) that now make up the Kalenjin, to bring together those

communities sharing common linguistic roots and inhabiting a particular area within

the Rift Valley and surrounding regions. The new identity was accepted by the

colonial authorities for greater ease of administration and with a view, perhaps, as

Lynch argues, of creating a bulwark against growing Luo and Kikuyu nationalism

(Lynch, 2007, Chapter 1). Political leaders from the communities now comprising

the Kalenjin gained by becoming part of a larger population group and a wider

support base. Similarly, there was no clear Luhya identity until the 1940s, when 17

separate communities sharing language and culture came together as the Luhya to

increase their ability to compete for land tenure rights and related benefits

(MacArthur, 2008, p. 230).

The important factor that needs emphasizing about the shifting alliances and the

political allegiances of Kenya‘s groups is that there is no long history of sustained

ethnic conflict between specific groups. Both before and during the early colonial

period there was coexistence between different communities, intermarriage and

trade (Klopp, pp. 487-9; Berman, 1990, p. 49). Conflict occurred over water, grazing

or cattle, but it happened alongside the other communal interactions noted above. It

was colonial occupation, the seizure of land and the creation by the British colonial

authorities of reserves for ―tribes‖ that created ―tribes‖ in specific areas with

potentially conflicting interests, landless peasants and developed a desire to reclaim

16

land taken from various communities. Land reclamation became a serious issue for

displaced communities and failure to address it the cause of serious and lasting

grievance for some of them.

Post-independence land resettlement schemes only enhanced the grievances of

non-Kikuyu communities. The perception became ingrained, especially among the

Kalenjin and Masaai, that ―outsiders‖ had gained at the expense of indigenous

communities creating a level of suspicion and political/community competition that

had not previously existed. This was fertile ground for politicians willing to incite

ethnic conflict for political advantage. Their ability to do this was enhanced,

according to Martin Gitau, the head of the Kenyan Journalists‘ Association, and

Dennis Itumbi, editor of the Fountain Post blog in Kenya, because as a general rule

Kenyans of one particular community will tend to vote en masse for the candidate

seen as representing their community in elections (interviews with Gitau and Itumbi,

10 February 2010, Nairobi).

The failure of Kenyatta and then Moi to deal with land ownership and related

problems meant that for many poor or landless Kenyans , the competitive elections

from 1992 onwards became opportunities to seek redress. This rendered poor, rural

Kenyans susceptible to politicians keen to exploit these issues to build local political

and electoral support. As the BBC‘s experienced Africa correspondent, Mark Doyle,

wrote in January 2008, he had observed that both ahead of and during the 2007

elections ―politicians from all ethnic groups. . . had been preparing the ground for

trouble in the wake of the elections because they know that ‗land clashes‘, as they

17

are known in Kenya, always flare up around polling time‖ and can be manipulated to

serve the interests of the politicians (Doyle, 2008).

Political parties were of less intrinsic importance in Kenya than in Western Europe or

the United States, for example, and generally served not as aggregators of

individuals with shared political beliefs or interests but as means for politicians to

build the machinery, resources, local, regional and national support to garner

sufficient votes to win power. Kenyan parties have always been shifting coalitions of

politicians seeking power and have been part of the alliances necessary for building

a national political presence. There was a slight difference in 2007 in that Odinga

was more clearly radical and populist than Kibaki and so issues rather than just

choosing between ―big men‖ and their alliances began to enter the picture. Odinga,

in the 2007 election campaign, was trying hard to appeal beyond ethnic community,

particularly to the poor and the young. Odinga appealed to young, poor Kenyans

who felt they had gained nothing from economic growth or decades of

independence, but who could see an elite in politics and business which had

benefited hugely; he had had massive support in his home province, Nyanza, and in

the Rift Valley, where Ruto and his allies had taken the support base that had

underpinned Moi for decades and built a Kalenjin following based on expectations of

a solution to land grievances and hopes of some form of federalism.

Kibaki‘s PNU was based on alliances more reminiscent of the days of Kenyatta, with

the Kikuyu at the centre, and support from smaller communities outside the Rift

Valley and Nyanza, along with the backing of declining politicians like Moi. The

incumbent president was viscerally opposed to Odinga‘s radical populism and was

18

unwilling to seriously contemplate a major redistribution of land or wealth or a

diminution of central control through federalism or a reduction in presidential powers.

The mix of competing approaches to major issues like land and the personal rivalries

combined with huge economic/social inequality and major grievances among key

communities to provide a wealth of combustible material that only needed a spark to

ignite substantial conflict. Behind it all was a deep well of frustration, anger and

deprivation among poor Kenyans of all communities—a well that politicians could

draw from to maintain or extend their own ambitions and privileges and damage

those of their political opponents. The grievances could be exploited to build votes,

intimidate opponents or fight an unwanted election result.

When the voting was over and initial results emerged from the electoral commission,

the ODM was ahead in the parliamentary vote and Raila Odinga had a marginal lead

that some opinion polls in Kenya suggested had stretched to a million votes. The

Kenyan Nation newspaper went so far as to report on 29 December that ―bar any

force majeure, Mr Rail Odinga is poised to win the presidency‖. But delays in

announcing the result of the presidential vote and rumours of fraud within the

electoral commission led to protests by Odinga supporters and a violent and lethal

reaction from the police. The clashes left several policemen and a larger number of

ODM supporters dead in the period between 27 and 30 December. The

announcement of the victory by Kibaki drew immediate accusations of cheating from

the ODM and expressions of concern over the veracity of the count from

independent observers. The initial violence was a direct result of anger over the

19

delays in the count and then over the result. The violence continued until the signing

of a political agreement by Kibaki and Odinga at the end of February.

Political violence with an ethnic face

Although it was the violence between ethnic groups which dominated reports of the

three months of crisis, it must be noted that in the opening days, the most serious

violence was from the Kenyan police and was directed against ODM demonstrators.

Attacks by supporters of one party from one community against perceived

supporters of their opponents among alien communities escalated and caused

hundreds of deaths, but police violence principally against ODM supporters was

constant (HRW, 2008, p.4). It became clear that during and after the election

campaign, leaders and party activists on both sides utilised local grievances, ethnic

stereotypes, insults and fear of ―others‖ to mobilise votes, boost attendance at rallies,

mobilise demonstrators and eventually involve people in intimidation and direct

violence against opponents. It is also clear that, in an atmosphere of distrust,

suspicion and fear, some violence was spontaneous and a certain gruesome tit-for-

tat pattern emerged in some areas beyond the direction of political leaders.

The violence took on an openly ethnic character, especially in the Rift Valley.

Kalenjin and Masaai gangs were utilised by ODM politicians and by local grandees

to attack their political opponents and to achieve through violence and threats the

destabilising of the government and, at least in the short-term, re-occupation of land

claimed as Kalenjin or Masaai. The violence on both sides was largely planned but

also developed its own spontaneous offshoots, as groups or communities reacted to

the spiralling crisis. The level of planning is likely to be revealed when the

20

International Criminal Court names those Kenyan political leaders who it wants to

indict for crimes relating to the violence. As David Anderson has pointed out, the

ODM was prepared before the election results were announced to cry foul and had

―laid plans for a campaign of civil unrest‖. This was to be so widespread as to make

it impossible for the government to cope with all the outbreaks across the country.

This campaign became violent partly due to the extremely violent response of the

police, but largely due to the deliberate cranking up of tensions by national and local

political leaders.

It is no surprise that ―the worst violence has occurred in areas where it is easy to

mobilise thuggery, such as the slums of Nairobi and in places where there is a long

history of animosity between neighbouring communities such as the resettlement

schemes of the Rift Valley‖ (Anderson, Prospect, 2008). But Anderson is clear in his

view that this animosity and the resulting ethnic violence is ―not rooted in any deep-

seated ethnic hatred, although no one would deny that as this crisis has mounted,

growing fear and latterly, a lust for vengeance has driven a wedge between

communities‖ (Anderson, Independent, 2008).

The reactions of politicians to the post-election violence did nothing to cool tempers

or discourage violence. Kikuyu and Kisii were targets in much of the Rift Valley.

Those targeted were seen as both political opponents and as interlopers who were

occupying land stolen from the ―rightful‖ owners. Kalenjin politicians in the ODM not

only used the lure of getting land back but also fear tactics that if Kibaki remained in

power, local communities would be under threat and would lose more land and come

under the control of outsiders like the Kikuyu. The demonstrations and violence

21

sparked off by the fraudulent election was used by these leaders to pursue long-term

underlying aims of their own.

In their in-depth report on the violence, Human Rights Watch pointed out that the

failure of successive Kenyan governments to address grievances had intensified

community animosity leading to serious ethnic divisions, that politicians who had

organised and funded political violence during previous elections had never been

brought to book, and concluded that, ―this violence is the outcome of decades of

political manipulation of ethnic tensions, and of impunity intertwined with

longstanding grievances over land, corruption, inequality and other issues‖ (HRW,

2008, pp. 2-3). The report added that the ODM had built a coalition ―based on the

widespread perception that the Kibaki government had entrenched tribalism and

governed in the interests of the Kikuyu community‖ (HRW, 2008, pp.4-5).

The organised violence against target communities was matched by substantial and

consistent vilification of opponents by political leaders on a national, regional and

local basis. Ethnic stereotypes were widely utilised to denigrate opponents. Raila

Odinga and his Luo supporters were repeatedly ridiculed for being just ―boys‖ – a

play on the different rituals for the progression from boyhood to manhood among

Kenya‘s different groups. The Kikuyu practise circumcision, the Luo don‘t. There

was racist, obscene and offensive campaigning on this issue by PNU leaders and

supporters. The circumcision motif had been used by Kibaki supporters during the

2005 referendum campaign and was used extensively to denigrate ODM politicians

and supporters in 2007 – even to the extent of calling for them to be forcibly

circumcised (KNCHR, 2006, p.45 and 2007; and Warungu, 2008). ODM politicians

and propaganda concentrated far more on land issues, alleged corruption, Kikuyu

tribalism and the dangers of Kikuyu hegemony to other groups. In such a situation,

22

the media was bound to become part of the problem even if they only reported

political rhetoric.

The media, political hate speech and the radio incitement

Elections are always a testing time for the media, even in societies with established,

varied and well-regulated media. In Kenya, political debate and elections take place

in a difficult and potentially hostile environment for the media and journalists.

During the period of one party rule under Kenyatta and Moi, there was limited

political debate and despite a variety of newspapers with diverse ownership there

was a strong authoritarian tradition which meant that while the Kenyan press was not

as directly censored or repressed as that in many African states, it served as an

instrument of the government with considerable self-censorship and a major

emphasis on the actions and pronouncements of the president and his senior

ministers (Heath, 1997, pp. 44-5). The state-funded public service radio and TV (the

Kenya Broadcasting Corporation - KBC) was subject to government control and

acted as the voice of the government, and, to a great extent, still does today.

As one-party rule came to an end, the self-censorship and authoritarian attitude of

government weakened but the opening up of the press and broadcasting was a slow

and far from easy process – as the author can testify from personal experience,

having been arrested, briefly detained and then deported from Kenya for trying to

enter the country in February 1991 to make a BBC radio documentary on the calls

for an end to one party rule.

After 1992 and the formal legalisation of a multiplicity of political parties, the press

(notably the Nation, owned by the Agha Khan, and the Standard, formerly owned by

23

the Lonrho multinational based in London, but now owned by group of Kenyan

businessmen and generally seen as an opposition newspaper—it is an open secret

in journalism circles in Kenya that the Moi family owns the Standard, while it is a

similarly open secret that controversial Kalenjin leader William Ruto is one of the

owners of Kass FM and that Uhuru Kenyatta owns K24 TV and Kameme FM radio)

became a forum for lively political debate (Heath, 1997, p 46). During the 2007-8

elections, it was considered that the press and broadcast media was relatively free to

reflect the political debates (EUEO, 2007, pp.1-2) as it had been during the 2005

referendum campaign and, to a lesser extent because of Moi‘s incumbency, in 2002.

But observers felt strongly that the media failed to provide equitable coverage of the

political leaders and parties. Internews, an American-based NGO working in Kenya

to help train journalists, believes that the picture is mixed and that extensive

coverage was given of the campaign by the print media, but that there were, as there

are for example in Britain during elections, clear preferences expressed or implied

for particular candidates or parties. There was also a clear need for better training

for journalists to cover the intricacies, controversies and politically sensitive issues

arising during elections (interviews with Rambaud and Jooste, Nairobi, 17 February

2010).

The state-owned Kenya Broadcasting Corporation was criticised heavily by EU

election monitors for failing ―to fulfil even its minimal legal obligation as a public

service broadcaster...its coverage demonstrating high degree of bias in favour of the

Party of National Unity (PNU) coalition‖ (EUEO, 2007, p.2). A study commissioned

by the United Nations Development Programme (UNDP) found that ―even though the

leading newspapers, television and radio stations were not very biased for or against

any of the candidates, there were discernable preferences shown by the tilt they

24

gave in favour of or against candidates and their campaign issues‖. The report also

noted that vernacular FM radio stations showed clear preferences for the candidates

and parties ―whom they perceived to be the choice of their listeners‖, (UNDP, 2008).

The behaviour of the vernacular stations gave rise to accusations (see in detail

below) that they were acting as hate radio broadcasters.

Overall, observers of the Kenyan media felt that the mainstream press and

broadcasting organisations were not inciting hatred or playing political games, but

that they failed to prevent the dissemination of party propaganda and the violent

rhetoric of many political leaders, and they failed ―to live up to professional and

ethical standards‖ (BBC WST, 2008, p.2). Some Kenyan journalists went further in

their criticism at a workshop in Nairobi in January 2008, accusing the print and

broadcast media of being willing to put money ahead of responsibility by ―accepting

and conveying paid-for hate material‖ which could have incited the audience at a

time of tension (IPS News, 2008).

The general thrust of comment on the media – leaving aside the vernacular stations

for the time being – was that while Kenya had a relatively free press by African

standards, there was a lack of professionalism and training that would enable

journalists to effectively cover political controversy, conflict and elections and a

vulnerability because of a lack of viable media laws and legal protection for

journalists. This rendered the media and journalists vulnerable to politicians wishing

to intimidate or manipulate them to get across party propaganda and to incite unrest

or violence during and after the campaign. Journalists and editors were also subject

to pressure from commercial interests (whether newspaper/radio owners or their

advertisers and financial backers, and , of course, from politicians who secretly

owned media outlets or owned them via proxy businessmen). The BBC World

25

Service Trust, in its report on the media and the elections, made clear the political

and commercial pressures and constraints that affected the independence and

integrity of the media and highlighted the need for the strengthening of the freedom

of the press, the removal of political and economic constraints and the creation of an

environment in which journalists could do their work in safety and free from pressure

or intimidation (BBC WST, 2008).

The level of interference by the government was obvious, too. The European Union

monitors explicitly criticised the ejection of all but a chosen few ―trusted‖ journalists

and organisations from the venue in which the hurried announcement of Kibaki‘s

victory was made and emphasised the lack of media freedom resulting from the

order by Kibaki‘s Internal Security Minister ―to suspend all live broadcasts, seriously

infringing the rights of the media to report without undue state interference‖ (EUEO,

p. 1). While the Kenyan government argued that the live ban was to prevent

broadcasting of film of violence that could provoke further conflict, the ban gave KBC

(under government control) a monopoly on TV and radio reporting.

As the violence increased this gave commercial and vernacular stations greater

credibility when they broadcast accounts of events or statements by political leaders

that contradicted the official position, which was widely believed to be inaccurate.

This led to a very skewed and chaotic media environment after the elections. The

commercial sector in broadcasting has low standards of editorial control, untrained

staff and little experience and, as Maina has cogently argued about the emerging

private sector, ―The private broadcasters, while seeming to take the duty of informing

the public seriously, exhibit a tendency towards bias, and almost every channel can

be identified with a political party or personality‖ (Maina, 2006, p. 9).

26

The UNDP report on the media is even more forthright about bias in reporting the

violence after the 2007 elections: ―The coverage of the post-election violence by the

media brought to the fore the entrenched ethnic divisions as various media houses

took obvious positions for against the status quo‖ (UNDP, 2008). As the violence

escalated from demonstrations about the results by the ODM and an extreme police

response to clearly orchestrated violence against specific targets with an

increasingly ethnic character (even though this paper argues that the causal factors

were political, social and economic rather than primarily ethnic), the partiality of the

media became part of the problem.

What this brief summary of the role of the mainstream press highlights is the

vulnerability of journalists and the media in the face of determined attempts by

politicians to see the media as weapons in their political battles. In the election and

referendum campaigns from 1992 onwards, politicians had no scruples about using

all available methods, including hate speech and the incitement of communal or

ethnic violence to advance their own interests.

The mongoose and the chickens - genocide and ethnic cleansing in

political rhetoric

Kenyan politicians have routinely used local grievances to set communities at each

other‘s throats and used language designed to denigrate and dehumanize their

opponents. In the period from independence to 1991, Kenya was a de facto and

later de jure one-party state in which the media was not free to report and in which

politicians in power had free rein to physically or verbally harass opponents or critics.

As the pressure built for multi-party politics, the language of political discourse

emanating first from Moi and his supporters within KANU and then spreading

27

throughout the competing political elites was one of insult, threat and accusations,

which ranged from corruption, through ethnic supremacism, to ethnic cleansing,

murder and genocide. Substantive policy issues or even the political or

administrative capabilities of individuals or parties were not at the heart of political

debate,— incendiary personal diatribes and the denigration of personalities, parties

and whole communities were.

In the early 1990s, Moi and his ministers used the rhetoric of fear to oppose multi-

partyism, claiming it would destroy unity and lead to fragmentation. They threatened

multi-party advocates and at public meetings called on KANU supporters to ―oppose

selfish troublemakers‖ and ―some regime supporters themselves appeared to

advocate violence against political dissidents, publicly urging citizens to cut off the

fingers of multi-party advocates, and to arm themselves with rungu (knobbed sticks)

and spears to crush opponents of one-party rule‖ (Haugerud, 1997, pp.76-77).

This political discourse of violent rhetoric and the dehumanization or denigration of

opponents became dominant during elections, periods of political tension and in land

disputes after 1992. Much of the language used is extreme and calls openly for

violence and killing, even though it is not clear that speakers realistically expected

their audiences to follow the instructions rather than treating them as symbolically

strong statements of contempt for opponents.

During the bitter and violent 2005 referendum campaign, a pro-Kibaki local council

leader in the Tatu area told his audience at a ―Yes‖ rally that, ―Raila [Odinga, the

―No‖ campaign leader] the monster should be hit on the head and killed so as not to

destabilize the Kibaki government‖; at another ―Yes‖ rally, Energy Minister and Kibaki

stalwart Simeon Nyachae demanded that ―those who are not circumcised [Luo like

28

Odinga] should be taken for circumcision ceremony‖; at a ―No‖ rally pro-Odinga MP

Joe Khaimi said that critics of Odinga should be lynched (KNHCR, 2005, pp. 30-33).

The circumcision insult was used frequently during 2005 and 2007 about Odinga and

his Luo supporters and during violence in both campaigns forcible circumcision of

Luo by Kibaki supporters was an horrific part of the political violence.

The Kenyan National Human Rights Commission (KNHCR, 2005) has reported fully

on the use of dehumanizing rhetoric and hate speech by politicians on both sides of

the political divide during the referendum. In a later report, Still Behaving Badly

(KNHCR, 2007), the organisation said that the 2007 election campaign had been

marked by continuing use by politicians of insults against opponents, threats of

violence and effective incitement to violence. The commission noted a slight

decrease in open use of such speech, but added that ―covert hate speech,

defamatory and unsavoury language continues unabated‖ and that ―unfortunately,

Kenyans continue to condone and cheer hate speech and have themselves become

active agents of proliferation of hate campaigns against politicians and fellow

Kenyans‖ (KNHCR, 2007).

Some of the discourse – as happened in Rwanda with the use of the term cockroach

or inyenzi for Tutsis and innocuous expressions like ―go to work‖ to mean killing

Tutsis – was seemingly unrelated to politics but was clearly understood by

protagonists. There were frequent references to the need for the "people of the milk"

to "cut grass" and complaints that the ―mongoose‖ has come and "stolen our

chicken". This, to Kenyans, is easily understandable with the pastoralist Kalenjin

referring to themselves as people of the milk, the ―grass‖ refers to non-Kalenjin or

non-Masaai in areas claimed as land originally belonging to them, and the mongoose

is a reference to Kikuyus who have bought land in Rift Valley and are viewed by the

29

Kalenjin and Masaai as interlopers and essentially thieves (AllAfrica.com, 2008).

The calls by pro-ODM leaders for its supporters to ―cut the grass‖ was open

incitement to attack what were viewed as foreigners in the Rift – cutting the grass

meant killing them or driving them from land claimed as Kalenjin.

Pro-Kibaki politicians didn‘t just use the references to circumcision but frequently

derided Odinga, the ODM and its supporters as ―beasts from the west", "baboons" or

"animals of the west" (AllAfrica.com, 2008). Although the comments on occasion

were accompanied by direct calls for violence they drew on cultural differences and

negative stereotypes and referred back to disputes about access to land, wealth and

control of state power. The human rights commission concluded in its report on the

2005 referendum that even where there wasn‘t a direct call for violence, this political

language and ―the resulting stigmatization, dehumanization and hatred is just as

harmful‖ (KNHCR, 2005, p.26).

The 2007 electoral campaign showed little substantial decline in violent and

offensive political rhetoric. Once the dispute over the result turned to violence and

that violence had developed an ethnic edge, the rhetoric of genocide and ethnic

cleansing came into play – but not calls for genocide, rather accusations that their

opponents were using carrying out genocide against supporters of the targeted

group. Hate rhetoric became a polemical weapon to be wielded against your

opponents, denigrating them not only in the eyes of Kenyans, but also internationally

and, in particular, in the eyes of the foreign dignitaries (Kofi Annan, African Union

mediator John Kufuor and US Assistant Secretary of State for African Affairs Jendayi

Frazer) involved in negotiating an end to the violence.

30

By early January, both sides in the conflict were using the term genocide to refer to

the policies and actions of their opponents. On 2 January, President Mwai Kibaki's

government accused rival Raila Odinga's party of unleashing "genocide" in Kenya as

the number killed in the violence passed 300 – "It is becoming clear that these well-

organized acts of genocide and ethnic cleansing were well planned, financed and

rehearsed by Orange Democratic Movement leaders prior to the general elections,"

according to a Kibaki government statement delivered by the Lands Minister, Kivutha

Kibwana (Reuters, 2008). The ODM wasn‘t slow in replying and it accused the

government of repressive policies "bordering on genocide" by ordering police to

shoot protesters demonstrating against Kibaki's victory (Reuters, Jan 2008).

A website was set up by Kibaki and PNU supporters entitled Chronicles of the

Kenyan Genocide, which sought to amass evidence from the PNU side that Kalenjin

and other pro-ODM ethnic groups were carrying out systematic genocide against

Kikuyu supporters of the PNU, naming specific ODM leaders and accusing them of

inciting. The site accuses an ODM MP of inciting Kalenjin and Luo youths to drive

Kikuyu and Kisii people from Molo district, to attack and kill them and that he and

other ODM officials provided transport to enable attacks to take place that led to the

killing of Kikuyu and the destruction of their homes (Chronicles of the Kenyan

Genocide, 2008).

Despite this, from the foregoing narrative of the conflict, it is clear that despite the

organised and coordinated nature of many of the attacks, there was no obvious

intention on the part of any group to annihilate another. The hate rhetoric was not

part of a campaign to rid Kenya of all members of any one or collection of identifiable

ethnic/linguistic groups. Taking Chalk and Jonassohn‘s definition of genocide as ―a

form of one-sided mass killing in which a state or other authority intends to destroy a

31

group as that group and membership in it are defined by the perpetrator‖, then what

happened in Kenya was not attempted let alone actual genocide (Chalk and

Jonassohn, 1990, p.23). The Kalenjin and their political allies showed no intention of

wiping out the Kikuyu and the Kikuyu were not powerless victims—they were

involved in tit-for-tat killings of Kalenjin, Luo and Masaai.

Like Chalk and Jonassohn, I would exclude from the definition of genocide ―those

cases of mass killing massacres, riots and so forth that had a lesser aim, no matter

how objectionable such cases are‖ (ibid). The term genocidal massacre also does

not readily apply to the Kenya situation as the actual disappearance of a group was

not intended (ibid, p.26). Kiernan uses genocidal massacres and also genocidal

moments to describe ―shorter, limited episodes of killing directed at a specific local or

regional community‖ undertaken to serve as object lessons for other members of the

targeted group and not necessarily intended as part of a campaign to destroy that

group (Kiernan, 2007, pp. 13-5).

In Kenya, the ―massacres‖ and ―moments‖ were clearly organised by political and

community leaders and had definite goals – creating a situation of ungovernability

that would force the Kibaki government to negotiate over the disputed election

results, to create a new political landscape in areas like the Rift or in PNU dominated

regions in which perceived opponents from other communities were driven out and

given a clear message not to return, and also the clearing of land claimed by one

community of what were seen as ―others‖ or interlopers. But even if a definition of

genocidal massacre or moment could be applied to particular events, those events

were not part of a wider campaign of genocide.

32

The reports of the human rights groups quoted so far indicate the organised and

planned nature of the violence. Some of the instigators have been remarkably candid

about their intentions. In an interview with the BBC‘s Pascale Hurter, Kalenjin

community and political leader Jackson Kibor openly advocated the killing of Kikuyu

and said, ―We will fight. This is war. We will start the war. We will divide Kenya‖ (BBC

World Service, 31 January 2008 and HRW, 2008, p.39). Human Rights Watch also

details threats made by Kalenjin and ODM leader and former Moi lieutenant William

Ruto against the Kikuyu and incitement to violence by him on the basis of Ruto‘s

belief that Kibaki was governing ―this country on the basis of tribalism‖ (Ibid).The

media was one medium for those advocating violence or disseminating hate speech

to get their messages across.

Vernacular radio and hate speech

―Radio is the premier means of reaching the public with news and information in

countries where most of the population is illiterate and television sets are rare‖

(Chalk, 2000, p. 93). Kenya has a higher rate of literacy than many of its neighbours

(above 70%, but with an uneven spread across age ranges, urban/rural divides and

gender). It also has an increasingly diverse network of newspapers, TV and online

news providers. But as Frank Chalk identified in relation to Rwanda, with limited

access of populations in Africa, particularly in rural areas, to TV let alone the Iinternet

and greater levels of illiteracy or partial literacy in the countryside, radio remains a

key source of news and other information. It is cheap, does not require reading skills

and is immediate and, with the growth of FM stations across Kenya, available with

strong local content and in vernacular languages.

33

Kenya does not have a long history of vernacular radio. Independent Kenya

inherited the colonial radio system – geared mainly towards the interests of

administration, economic development and settler interests. The Kenyatta and Moi

governments retained close control over radio and emphasis was put on national

unity and to broadcasting in KiSwahili and English rather than vernacular languages.

It was the end of single party rule and the gradual opening up of the media to greater

freedom of expression which led to pressures for local, vernacular radio. Fearing its

use for political purposes by the opposition and the loss of control of a key means of

influencing opinion, Moi initially opposed granting FM licences to commercial stations

– particularly as the first request for a licence to broadcast came from Royal Media

Services owned by S.K. Marcharia, a prominent businessman with close links to

leading opposition politicians. He wanted to set up a Kikuyu FM station, Kameme

FM. Despite Moi‘s opposition, pressure on the government led to the licensing of

commercial FM stations in 1996 (Wafula, 2008). But the government was not happy

with the situation and in 2000 banned Kameme on the grounds that it was being

used to campaign for the opposition – exactly what Moi was using KBC for on behalf

of KANU. But realising the power of vernacular broadcasting, the government

launched another Kikuyu station, Inooro FM, to compete with Kameme, which was

eventually allowed to broadcast again. Both Joshua arap Sang, chief of operations

and lead presenter for Kass FM, the Kalenjin station, and Macharia Wamugi,

operations head for Kameme FM, a popular Gikuyu station, told the author in

interviews in February 2010 that the governments of Moi and then Kibaki had been

suspicious of vernacular radio because large parts of the population would not be

34

part of the dialogue between local stations and the audience in their own languages

and that they thought this would be both divisive and beyond their control.

Following KANU‘s electoral defeat in 2002, there was a rapid expansion in FM

stations – particularly those broadcasting in Kikuyu, Kalenjin, Luo and Luhya. Kass

FM became the most influential Kalenjin station, while Lake Victoria FM and Ramogi

FM were the leading Luo stations. They were criticised by human rights groups

during the 2005 referendum campaign for inciting political violence – Inooro FM was

pro-Kibaki and broadcast songs deriding ―beasts from the West‖, meaning Odinga

and his supporters (KNHCR, 2007). The Kibaki government briefly suspended Kass

FM in November 2005, accusing it of inciting violence during the referendum

campaign (Wafula, 2008) – Odinga supporters accused the government of attacking

Kass because it was independent from the government and broadcast the views of

Odinga supporters opposed to the planned new constitution. Kass was allowed to

resume broadcasting when it produced transcripts of programmes and succeeded in

proving that no hate messages had been broadcast. This gave the station greater

credibility among Kalenjin listeners and increased suspicion of the Kibaki

government.

In a survey of Kenyan broadcasting, the BBC referred to reports of the broadcasting

of hate speech by a number of vernacular radio stations in 2005 and to continuing

fears that vernacular stations ―could influence ethnic tensions‖ (BBC, 2009). These

fears appeared to be realised with the publication of a slew of reports in early 2008

that vernacular radio stations were playing a negative role in the violence following

the elections. The journalist and media commentator Evans Wafula sounded an

alarming and ominous note when he wrote that ―reminiscent of the notorious RTLM

in Rwanda, the media in Kenya is partly to blame for the post-election bloodshed in

35

Kenya. There are worrying echoes of a planned genocide being incited by local

radio stations that urged people to ‗arm themselves‘ against their enemies‖ (Wafula,

2008). The BBC reported on 14 February that the government had ordered an

investigation into claims that vernacular radio stations had engaged in hate

broadcasting during and after the elections (BBC, News, 2008).

Unfortunately, there are few transcripts available of the vernacular radio broadcasts.

The stations involved are small and do not keep large archives, while major

monitoring organisations (notably the BBC Monitoring Service, which has a

monitoring station at Karen, on the outskirts of Nairobi) were not monitoring

vernacular radio, only Kenyan broadcasters using English or KiSwahili (Greenway,

2009;Mundi, 2010). There was some monitoring of press and broadcasting by the

Steadman Group and Strategic Public Relations Research Ltd on behalf of human

rights groups and the UNDP, but even they have relatively few transcripts.

The Kenyan Human Rights Commission believes there is cause for concern over the

language broadcast by some of the stations and there is evidence that in the past

the vernacular stations (Kass FM, Lake Victoria, Kameme and Inooro are named)

have been responsible for ―spinning information to support candidates and parties

who are of the same tribe as their audience while openly castigating those who are

not of the same tribes‖ (KHRC, 2008) – evidence perhaps of bias, but hardly a

convincing argument that stations were adopting practices similar to RTLM in

Rwanda.

But the UN-linked IRIN news agency reported that ―inflammatory statements and

songs broadcast on vernacular radio stations . . . all contributed to post-election

violence‖ and warned that behaviour of vernacular stations was worrying given the

36

role of RTLM in Rwanda. The agency cited Caesar Handa of Strategic Public

Relations Research Ltd, who carried out media monitoring for UNDP, as saying that

―there‘s been a lot of hate speech, sometimes thinly-veiled. The vernacular radio

stations have perfected the art‖. Those singled out in the IRIN report were Kass,

Kameme, Inooro and Lake Victoria. Handa told the agency (and also supplied the

author with further information in e-mails) that talk and phone-in shows were the

worst and that callers or politicians/local leaders interviewed on the stations engaged

in incitement against other communities, which the radio presenters seemed

powerless to prevent or control. IRIN said that Kass FM repeated the hate speech

used by Kalenjin politicians against the Kikuyu, notably the warnings that the

mongoose ―has stolen our chickens‖ and that the ―people of the milk‖ had to ―cut the

grass‖ and were ―getting rid of the weeds‖(IRIN, 2008; Handa, 2009). What is not

detailed is whether Kass journalists themselves made these comments or whether

they were from contributors via phone-ins, e-mails, SMS messages or in interviews

with politicians. It is certainly the case that phone-in programmes and talk shows

were repeatedly identified as the main problem areas whether with political or

community leaders using them to incite violence or just ordinary Kenyans phoning in

and giving vent to their fears or prejudices. Joshua arap Sang, who presented

programmes on Kass FM during the elections and the violence, denies that his

station broadcast hate speech; he said they were just broadcasting to their people

(the Kalenjin) and explaining the situation to them in language they would

understand. When presented with transcripts detailing incitement, for example, to

beat opponents he denied they had been broadcast. Despite this, there is evidence,

both in the few transcripts and in the recollections of media monitors and local

37

journalists, of regular use of ethnically demeaning language and threatening

language by some presenters

Western journalists covering the violence in 2008, also drew attention to the content

of vernacular broadcasts. Mike Pflanz, reporting for the British Daily Telegraph and

the Irish Independent, said ―there is growing evidence that hate-filled broadcasts

have poured fuel on the fire of Kenya‘s post-election killings and contributed to

―ethnic-cleansing‖ in certain areas‖. He said this was ―a chilly echo‖ of Rwanda and

the role of hate radio in the genocide. He went on to say that programmes and

songs broadcast on the vernacular stations ―had helped incite tribal killings‖ and he

quoted Kamanda Mucheke of the Kenyan National Commission on Human Rights as

saying, ―it has been thinly-veiled, but it is clearly hate speech and to a large extent

the violence we‘re seeing can now be attributed to that‖ (Pflanz, 2008).

What is lacking from all the reports so far about the role of hate radio is clear

evidence of the content that is said to have had such an effect in inciting violence.

Pflanz does not cite the content of a single broadcast and transcripts are lacking for

references in the human rights report. Of 13 transcripts supplied to the author by

Caesar Handa, only one, from Kass FM on 3 October 2007, in the run up to the

elections, has any incitement to violence and that could be as much symbolic as real

incitement to cause harm. The offending broadcast had a presenter reading out the

following SMS from a listener – ―Leaders who abuse ought to be shown, they should

not be elected at all or they be beaten and their property be burnt‖ (Handa, e-mails,

2009). A broader survey sent to the author by Mr Handa indicates that strongly

partisan content was broadcast by Kass, Kameme and Inooro, but little that could be

compared with RTLM‘s output in 1993 and 1994, which set the agenda for the

38

genocide against the Tutsi – representing them as blood enemies who had to be

wiped out to prevent the mass killing of Hutus.

Greater credence is given to the accusations of hate broadcasting by East African

media organisations and by Kenyan journalists themselves, including some working

for FM stations. The Tanzanian-based IPP media organisation looked at the

monitoring information available and at comments by Caesar Handa and concluded

that the Kass, Kameme, Inooro and Lake Victoria FM stations had been the worst

offenders in spreading hate messages, especially through talk shows with guests

and phone-ins (IPP, 2008). Another local journalist, Dennis Itumbi, said that all the

above stations had been broadcasting hate messages and that a journalist

broadcasting for Kass had called on his Kalenjin audience to ―leave your houses,

war has begun‖ and to ―arm themselves‖. Itrumbi also says that a Reverend Kosgey

who broadcast on Kass was active in organising attacks on Kikuyu communities in

the Rift Valley. He also says that the Kikuyu station Inooro ―was particularly blamed

for organising revenge attacks in Kenya‘s Central Province‖ (Itumbi, 2008 –

confirmed in conversation with the author in Nairobi, 10 February 2010).

The viewpoint that hate broadcasting was being carried out is supported by

journalists who attended an Inter-Press workshop in Nairobi in January 2008 to

discuss the role of the media in Kenya‘s crisis. The reports of the meeting quote a

number of Kenyan journalists lamenting the failures of journalists and, worse, the

role of some journalists in perpetrating hate broadcasting. David Ochami of the

Media Council of Kenya says that from long before the elections, the vernacular

radio stations had served to ignite ―ethnic consciousness‖ among their listeners,

making them ―support leaders from their own tribe and harbour bad feelings about

people from other communities‖. Ochami cites an unnamed journalist working in

39

vernacular radio as disclosing that ―the ethnic hate our radio station was propagating

about those from outside the community was unbelievable. I can‘t repeat any of

those expressions at this forum‖. The journalist went on to say, though, that the

expressions of hate came largely from calls to phone-ins, but ―the unfortunate thing

is we let these callers speak vile and then laughed about [them]‖. The report of the

workshop cites another anonymous journalist as admitting that ―we took sides in the

issue and we became subjective, forgetting our professional tenet of objectivity and

neutrality. In fact, this polarisation was so bad in the newsrooms that some

broadcast journalists refused to cover or read news that wasn‘t favourable to the

candidate or the party they supported.‖

Other participants said that broadcast and print media were too ready to accept

money to carry campaign messages and they put money ahead of responsibility by

―accepting and conveying hate material‖. Some also spoke of objectivity almost

always giving way to partisanship in reporting. Some blamed media station owners

who had ―vested interests in either camp of the political divide‖ (all quotes from IPP,

2008).

Because of the lack of substantial transcript material or recordings it is impossible to

estimate, let alone measure accurately, the time period over which vernacular

stations carried hate messages, the proportion of air time they took up, the role of

journalists in directly inciting hatred and violence and the extent to which journalists

through insufficient training, experience and editorial direction became caught up in

partisan reporting and allowed, almost by default, hate speech to be broadcast.

Using the analytical tools from discourse analysis and from the study of hate radio in

Rwanda is difficult with the dearth of verifiable broadcast material. However, from

40

the little available and the reports of observers and journalists from the stations

quoted above, there is at least an opportunity to look at the general agenda setting,

framing and representation of subjects within vernacular broadcasts.

Kass FM is most often accused of partisanship and broadcasting hate messages in

2005 and again in 2007-8. Along with Kameme and Inooro and, to a lesser extent,

Lake Victoria FM, it seems to have followed a general editorial line of favouring

candidates from its own community, broadcasting material that favoured local

Kalenjin candidates supporting the ―No‖ campaign and the ODM, and denigrating

those, generally Kikuyu or Kisii, who supported the ―Yes‖ campaign and then the

PNU. Strongly derogatory terminology was used and questionable calls to ―cut the

grass‖ and get rid of ―weeds‖ – generally accepted as shorthand for clearing

outsiders from what was considered Kalenjin land. But the frequency of such

broadcasts, the proof of their origin (whether Kass journalists or interviewees/callers)

and whether they were exceptions to the rule or part of a routine pattern of partisan

broadcasting impossible to assess accurately. The same goes for the other stations

– the Kikuyu ones may have referred to ―beasts‖ or ―animals‖ from the West and to

issues around circumcision but how often? In what context? With what level of

incitement to hatred or violence?

What we can conclude is that there is evidence of them having a partisan agenda.

They clearly framed references according to this agenda – local candidates were

supported and praised while the representation of opponents was couched in

inflammatory language. On occasions, as some journalists have admitted, news

about other parties or candidates was omitted and some journalists became very

biased in their approach to reporting certain kinds of news. This certainly suggests a

discourse on these stations that was partial, open to the charge of inciting contempt

41

for others and lacking any clear standards of impartiality, balance and responsible

journalism. Joshua arap Sang of Kass FM, admitted in an interview that his station

was outspoken and was in touch with what it knew its audience wanted to hear. He

said he and other presenters spoke in the language the audience would understand

and appreciate. He effectively confirmed that they had strong views about particular

parties, politicians and issues and broadcast what they wanted to say in the way they

wanted to say it (Sang interview, Nairobi, 12 February 2010).

Hate Radio or Purveyors of Hate Speech?

Reviewing the known output and reported behaviour, how does it measure against

the role of RTLM in Rwanda? There is a large body (see for example Chalk, 2000;

Thompson, 2007; Strauss, 2007; Chretien; Des Forges; Melvern, 2000) of analysis

of hate radio and its role in the Rwandan genocide. Three Rwandans (Hassan

Ngeze, Jean-Bosco Barayagwiza and Ferdinand Nahimana) involved with RTLM and

with the anti-Tutsi newspaper Kangura were found guilty by the International

Criminal Tribunal for Rwanda of genocide, incitement to genocide and crimes

against humanity for their roles in broadcasting and publishing hate material that

directly incited hatred and called for the killings of Tutsis as a group and specific

individuals or collections of individuals. They were the first journalists since Julius

Streicher, editor of the violently anti-semitic Nazi newspaper Der Stürmer (who was

sentenced to death for incitement to murder and extermination of Jews, constituting

―a crime against humanity‖ at Nuremburg, but incitement to commit genocide in

today‘s terms) to be found guilty of genocide through their work as journalists.

The testimony at their trial and the trial of a Belgian journalist, Georges Ruggiu, who

worked for RTLM and was sentenced to 12 years imprisonment for inciting violence,

42

demonstrated a continuous, organised and carefully targeted stream of broadcasting

that set a long-term agenda of hatred and ultimate destruction of the Tutsi. RTLM

identified victims and gave instructions to Hutu to kill their neighbours and

compatriots simply because they were Tutsi. The ICTR judgement found that RTLM

―engaged in ethnic stereotyping in a manner that promoted contempt and hatred for

the Tutsi population. RTLM broadcasts called on listeners to seek out and take up

arms against the enemy . . . these broadcasts called explicitly for the extermination

of the Tutsi ethnic group‖. (ICTR judgement in Thompson, 2007, p.283).

Kenyan vernacular radios, as far as can be discerned from limited broadcast, printed

and anecdotal evidence and material from the author‘s interviews in Kenya,

periodically broadcast hate speech about perceived opponents from other

communities, at times appeared to condone or even incite violence or the expulsion

of people from particular areas and did demonstrate considerable partisanship.

However, unlike RTLM there is no compelling evidence of a coordinated campaign,

of an organised, long-term setting of an agenda of ethnic attacks let alone

extermination or very clear organisational links (as the ICTR established between

RTLM and Kangura and the Hutu military and political hierarchy which organised the

genocide) between journalists, radio stations and those political leaders directly

involved in inciting and organising violence.

Clearly, many journalists became partisan, gave in to pressure from owners and

local/national political leaders and either through this partisanship or through fear or

inexperience allowed hate messages to be broadcast during talk shows, interviews

or phone-ins. In a lot of cases, it was sheer inexperience in hosting phone-ins or talk

shows that meant that by default rather than design hate speech or incitement (some

planned by politicians, but some the spontaneous contribution of ordinary citizens)

43

was broadcast. In some cases, the stations had clear political agendas and these

then could easily, by design of owners or politicians or in the heat of the moment, slip

over from partisanship into incitement.

But there is no evidence that would allow a direct comparison with Rwanda. What

there is evidence of, as identified by studies by Maina, the UNDP, BBC World

Service Trust and others, is clear partisanship on the part of the FM stations

mentioned in this paper, poor editorial standards and a willingness to become

―purveyors of the numerous rumours that circulated‖. They perpetuated divisions

and images of ―those who were perceived to be the aggressors and those who were

aggrieved (UNDP, 2008). The stations allowed themselves at times to be used by

powerful political and community groups with their own agendas. Kass, according to

Human Rights Watch, was not proved to be responsible for having a policy of hate

speech, but it allowed guest speakers and callers to express hatred and engage in

incitement against targeted groups without hindrance (HRW, 2008, p. 36). The

station, as demonstrated during my visit there, clearly has a ―Kalenjin‖ agenda

relating to issues like land and this leads to strongly partisan output and a clear

editorial agenda, but that alone is not proof of concerted, organised hate

broadcasting.

There was a lack of awareness of ethical issues and the broadcasting patterns of the

FM stations identified, as Maina had pointed out in 2006 after the prevalence of the

use of hate speech during the 2005 referendum campaign, an urgent need ―to

protect the rights of the audience, uphold professional reporting and language,

respect African culture and desist from ethnic and chauvinistic tendencies‖ (Maina,

2006, p.73).

44

The BBC World Service Trust reached similar conclusions, noting that vernacular

stations had incited fear and hatred, had operated without clear ethical standards

and had been ―routinely partisan‖. The Trust noted the particular problems with talk

and phone-in shows ―which have provided the greatest opportunities for hate

speech‖ and urged better training for journalists and presenters, clearer professional

and ethical standards and better monitoring of the media to identify problem areas

before they become involved in ―fanning ethnic hatred and fuelling violence‖ (BBC

World Service Trust, April 2008, pp.2-3).

To answer the question at the beginning of this section, Kenyan vernacular radio

stations clearly played a role in worsening tensions, exacerbating fear and suspicion,

perpetuating negative stereotypes and at times inciting violence. They allowed or

perhaps failed to prevent themselves being manipulated and allowed the relatively

unrestricted expression of hatred on phone-ins and talk shows. But unlike RTLM in

Rwanda, the stations did not do so as deliberate and sustained acts of policy. They

were partisan and allowed their stations to be used as vehicles for spreading

messages of hate by politicians and community leaders. But they played no

discernable role in setting an agenda for the extermination of groups and once

monitoring had revealed their activities and opportunities for thought and discussion

emerged (as with the IPP workshop) there were positive developments and ―local

language stations did also play a reconciliatory role – Kass did broadcast callers

calling for peace and for cooperation with human rights groups‖ (BBC World Service

Trust, April 2008 p.5).

The Kenyan media works in a political and economic environment that exerts great

pressures on them to be partisan and to bow to pressure to broadcast statements

from national or local leaders: there is a lack of laws or regulatory instruments which

45

protect journalists, provide legal safeguards for freedom of speech, identify and

provide means of combating hate speech or set basic standards of professional

journalism. And there is insufficient provision of training to ensure the capability to

meet and maintain those standards (see BBC World Service Trust, April 2008; BBC

World Service Trust, 2008; UNDP, 2008; and Maina, 2006).

The wider political environment is one that is dominated by frequent resort to the

incitement of hatred, denigration of opponents and a willingness to use violence and

the manipulation of community grievances as means to desired political ends. It was

clear in 2007-8, as in 2005 and the Rift Valley violence of the 1990s, that, as the

official report on the violence found, there ―were systematic attacks on Kenyans

based on their ethnicity and political leanings. Attackers organised along ethnic lines,

assembled considerable logistical means... ―(CIPEV, 2008, p.viii); people were

mobilised by political leaders manipulating the ―feeling among certain ethnic groups

of historical marginalisation, arising from perceived inequities...This feeling has been

tapped by politicians to articulate grievances...This has created an underlying climate

of tension and hate‖ (CIPEV, 2008, p.23). The grievances of some groups and the

generation among other groups of a fear of losing access to land, resources or the

advantages of being supporters of the governing elite, dominated political discourse,

while the use of armed gangs of party supporters or hired thugs became an

everyday occurrence during elections, referenda or in land and other disputes. This

made ―violence the method of choice to resolve a range of political differences and to

obtain political power‖, if other more legitimate methods did not suffice (CIPEV,

2008, p.35). It is hardly surprising that the media became another instrument in the

all out battle for political power and journalists were poorly positioned to resist.

46

The violence in Kenya was not attempted genocide and the behaviour of the media

did not equate either qualitatively or quantitatively with RTLM in Rwanda. But there

is considerable evidence that the political culture still dominant in Kenya is one in

which all means are viewed as legitimate in the winner-takes-all approach to politics

and the distribution of economic benefits.

A Chronology of Events:

2007

Early December 2007 - Sporadic violence occurs during the election

campaign, with the harassment of party supporters, officials and candidates in

heartland areas of opponents. Opinion polls show Raila Odinga‘s Orange

Democratic Movement (ODM) ahead of President Kibaki‘s Party of National

Unity (PNU).

27 December : Voting begins in parliamentary and presidential elections

28-30 December - Odinga‘s ODM ahead in parliamentary vote and Odinga

believed to be edging ahead in presidential race. Outbreaks of violent

protests by ODM supporters amid rumours of electoral fraud in Kibaki‘s

favour.

30 December - The Electoral Commission halts the count and then declares

Kibaki to be the winner of the presidential election; he is sworn in with haste.

The government blocks live TV news broadcasts.

31 December - ODM supporters start street protests and clash with police –

violence ensues with a harsh police reaction, a ban on live TV broadcasts and

heavy security force presence in areas with high-levels of ODM support.

47

2008

1 January - Increasing scale of protest, police reaction and violence between

political factions, with growing numbers of attacks in Rift Valley by ODM

supporters on perceived PNU voters. In Eldoret, Rift Valley, at least 30 Kikuyu

fleeing violence and intimidation die after church in which they were sheltering

from ODM supporters is burned down.

4 January - Kibaki says he will accept a re-run of the election only if the

Kenyan High Court orders it

2 January - Government ministers accuse Odinga supporters of ―ethnic

cleansing‖ and genocide in Rift Valley.

5-11 January - Attempted mediation by US envoy, African Union chairman,

Kofi Annan and others. Little progress is made and ODM calls for

international sanctions to force Kibaki to annul the elections. Violence

continues in Nairobi slums, Rift Valley, Kisumu and other areas as ODM and

PNU supporters attack each other and security forces react with lethal force to

opposition protests. Further accusations by both sides or ethnic cleansing

and genocide

8 January - Kibaki appoints new cabinet; further violent protests in Odinga

stronghold of Kisumu.

Mid-January-28 February - Violence continues, concentrated in Rift Valley

and Nairobi. Growing evidence is uncovered by human rights researchers of

planning and organisation of attacks on rival groups by ODM politicians in the

Rift Valley and by Kibaki supporters and the Kikuyu-based criminal gang, the

Mungiki. Violence gradually dies down as Odinga and Kibaki agree a power-

sharing deal.

48

Between 1200 and 1500 people died and between 300,000 and 500,000 were

displaced during the violence – at least 450 of the dead killed by the police.

(Compiled from BBC News Online, Guardian, Independent, Telegraph and

Human Rights Watch reports).

Bibliography

AllAfrica.com.2008. ―Kenya: Spreading the Word of Hate.‖ 22 January 2008.

http://allafrica.com/stories/200801220692.html)

Allen, Tim and Seaton, Jean.1990. ―Introduction‖, in Tim Allen and Jean Seaton

(eds), The Media of Conflict: War Reporting and Representations of Ethnic Violence,

London: Zed Books, pp. 1-7.

Anderson, David.(2008. ―Kenya‘s Agony‖. Royal African Society.

http://www.royalafricansociety.org/index.php?option=com_content&task=view&id=44

3 (accessed March 22, 2009).

Anderson, David.2008. Kenya on the Brink. Prospect, 142, January 2008.

Anderson, David. 2008. ―How Violence Affected Kenya‘s Democracy‖. Independent

30 January 2008.

BBC News.2009. Kenya media welcomes review of media law.

http://www.news.bbc.co.uk/1/hi/world/Africa/7817567.stm (accessed July 6 ,2009).

BBC News.2008. http://www.news.bbc.co..uk/1/hi/world/Africa/724531/stm

(accessed July 6, 2009).BBC World Service Trust. 2008. The Kenyan 2007

Elections and their Aftermath: The Role of Media and Communications.

49

http://downloads.bbc.co.uk/worldservice/trust/pdf/kenya_policy_briefing_08.pdf

(accessed July 7, 2009).

BBC World Service Trust. April 2008. Policy Recommendations, London, BBC

Berman, Bruce.1990. Control and Crisis in Colonial Kenya. London: James Currey.

Branch, Daniel and Cheeseman, Nic.2009., ―Democratization, sequencing and state

failure in Africa: Lessons from Kenya‖, African Affairs 108, no. 430:: 1-26.

Chalk, Frank.2000. ―Hate Radio in Rwanda,‖ in Howard Adelman and Astri Suhrke,

The Path of a Genocide: The Rwanda Crisis from Uganda to Zaire, New Brunswick,

NJ: Transaction Publishers.

Chalk, Frank, and Jonassohn, Kurt. 1990., The History and Sociology of Genocide,

New Haven and London, Yale University Press.

Cheeseman, Nic (2008), ―The Kenyan Elections of 2007: An Introduction,‖ Journal of

Eastern African Studies, 2, no.2 (July 2008): 166-84.

Chretien, Jean-Pierre. ―RTLM propaganda: the democratic alibi‖, The International

Development Research Centre.

http://www.idrc.ca/rwandagenocide/ev-108180-ido_TOPIC.html (accessed May 13,

20009).

Chronicles of the Kenyan Genocide. http://kenyangenocide.blogspot.com/

(accessed May 14, 2009).

50

CIPEV. 2008. Report of the Commission of Inquiry Into Post-Election Violence.

http:///www.khrc.org.ke/documents/Waki_Report_08.pdf (accessed April 1, 2009).

Des Forges, Alison.‖ Call to genocide: radio in Rwanda, The International

Development Research Centre.‖ http://www.idrc.ca/rwandagenocide/ev-108178-1-

do_TOPIC.html (accessed April 2, 2009).

Doyle, Mark.2008. Kenya‘s geographic and political rift, BBC News 28 January 2008.

European Union Election Observation Mission (EUEO) . 2008. Kenya General

Election, 27 December 2007, Preliminary Statement

http://www.eueokenya.org/MAIN/English/PDF/Kenya_2007

Fairclough, Norman.2003. Analysing Discourse. Oxford: Routledge.

Heath, Carla W. 1997. ―Communications and Press Freedom in Kenya,‖ in Festus

Eribo, and William Jong-Ebot. Press Freedom and Communication in Africa.

Trenton, N.J:Africa World Press.

Handa, Caesar.2009. E-mail to the author, July 1, 2009.

Haugerud, Angelique. 1997. The Culture of Politics and Modern Kenya. Cambridge:

Cambridge University Press.

Human Rights Watch. 2008. Kenya: Ballots to Bullets,

www.hrw/en/reports/2008/03/16/ballots-bullets (accessed May 10, 2009).

,Rodrigues, Simon. ―Corruption Trumps Tribalism,‖ International Herald Tribune

(IHT), January 10, 2008.

51

IPP.‖ Spreading the word of hate in Kenya.‖

http://www.kuryangu.com/ipp/guardian/2008/01/26/107059. (accessed May 24,

2009). Kwamboko Oyara. 2008. ―KENYA The Media is Not Innocent.‖ IPS News.

http://www.ipsnews.net/news.asp?idnews=41049 (accessed May 12, 2009).IRIN. ,

―KENYA: Spreading the word of hate.‖ January 27, 2008.

http://www.irinnews.org/Report.aspx?ReportId=76346 - IRIN (Integrated Regional

Information Networks) is part of the UN Office for the Coordination of Humanitarian

Affairs) (accessed February 23, 2009).

Itumbi, Dennis.―Kenyan radio stations criticised in human rights report.‖

http://blogs.journalism.co.uk/editors/2008/08/20/journalism_in_Africa_Kenyan_radio_

stations (accessed March 21, 2009).

Kenya Human Rights Commission. ―General Elections 2007: A Preliminary

Assessment.”

http://www.khrc.or.ke/documents/GENERAL%20ELECTIONS%2007_PRELIMINAR

Y%ASSESSMENT.pdf (accessed April 28, 2009).

Kenya National Commission on Human Rights.2005. Referendum Report,

http://www.knchr.org/dmdocuments/referendum.pdf (accessed April 28, 2009).

Kenya National Commission on Human Rights . Still Behaving Badly, December

2007. http://www.knchr.org/dmdocuments/Election_Report.pdf

(accessed March 23, 2009).

Kiernan, Ben.2007. A World History of Genocide and Extermination from Sparta to

Darfur, New Haven: Yale University Press.

52

Klopp, Jacqueline M. ―Ethnic Clashes and Winning Elections: The Case of Kenya‘s

Electoral Despotism.‖ Canadian Journal of African Studies, 35, no. 3 :473-517.

Kriegler Commission, Report of the Independent Review Commission, September

2008. http://www.hackenya.org/index (accessed June 27, 2009).

Kellow, Christine L, and Leslie H. Steeves, (1998), ―The Role of Radio in the

Rwandan Genocide,‖ Journal of Communications, 48, 3:109-128.

Lonsdale, John. 2008. ―Soil, Work and Civilization, and Citizenship in Kenya,‖

Journal of Eastern African Studies. 2, 2, July 2008: 305-14.

Lynch, Gabrielle.2008. ―Courting the Kalenjin: The Failure of Dynasticism and the

Strength of the ODM Wave in Kenya‘s Rift Valley Province,‖ African Affairs, 107,

no.429: 541-568.

MacArthur, Julie.2008. ―How the West was Won,‖ Journal of Eastern African Studies,

2, no. 2 (July 2008): 227-241.

Maina, Lucy W.2006. Kenya Research Findings and Conclusions. London: BBC

World Service Trust.

Melvern, Linda.2000. A People Betrayed: The Role of the West in Rwanda’s

Genocide, London: Zed.

Mueller, Susanne D..2008. ―The Political Economy of Kenya‘s Crisis,‖ Journal of

Eastern African Studies, 2, no. 2 (July 2008): 185-210.

Ndunde,Hezron.2008. ―From Cyberspace to the Public: Rumor, Gossip and

Hearsay in the Paradoxes of the 2007 General Election in Kenya.‖ Papers from the

53

conference of the Council for the development of Social Science Research in Africa,

December 7-11, 2008, Yaounde, Cameroon.

http://www.codesria.org/Links/conferences/general_assembly12/papers/hezron_ndu

nde.pdf (accessed May, 13, 2009).

Pflanz, Mike. 2008. ―Hate radio spreads a new wave of violence in Kenya.‖ Irish

Independent, January 28,2008.

http:// www.independent.ie/world-news/africa/ (accessed June 25, 2008).

Richardson, John E.2007. Analysing Newspapers: An Approach from Critical

Discourse Analysis. London: London.

Reuters . ―Kenya govt denounces "genocide" as toll hits 300.‖

http://www.reuters.com/article/topNews/idUSL277107920080102?feedType=RSS&fe

edName=topNews (accessed June15, 2008).

Somerville, Keith.2009. ―British Media Coverage of the Post-Election Violence in

Kenya 2007-8.‖ Journal of Eastern African Studies, 3, no 3: 526-542.

Strauss, Scott .2007. ―What is the Relationship between Hate Radio and Violence?

Rethinking Rwanda‘s radio Machete.,‖ Politics Society, 2007, no. 35: 609-637.

Thompson, Allan.2007. The Media and the Rwandan Genocide, London: Pluto.

United Nations Development Programme. (2008). Media Monitoring:The Experience

and Future, Strategic Public Relations Research Ltd, Nairobi.

Wafula, Evans.2008. ―I heard them cry for help as the media watched.‖

http://www.africanews/site/ I_heard_them_cry_for_help_as_the_media_watched

(accessed July 1, 2008).

54

Warungu, Joseph., ―Kenya in Crisis.‖ BBC College of Journalism Seminar. London:

Television Centre, February 13, 2008.

Interviews carried out by the author in Nairobi, Kenya, February 2010:

Martin Gitau, head of the Kenyan Journalists‘ Association, February 10, 2010.

Chris Greenway, interview with author, BBC Caversham, April 27, 2009.

Dennis Itumbi, journalist and editor of the Fountain Post blog, February 10, 2010.

Matthias Mundi, chief editor, BBC Monitoring, east Africcan Unit, Karen, February

11, 2010.

Moses Rono, former Kass FM journalist, February 11, 2010.

Joshua arap Sang, Head of Operations and presenter, Kass FM, February 12, 2010.

Macharia Wamugi, Operations Editor, Kameme FM, February 15, 2010.

Alfred Mutua, Chief Government Press Secrretary, Office of Public Communications,

February 16, 2010.

Ida Jooste, Internews Kenya Resident Advisor, February 17, 2010.

Brice Rambaud, Internews Kenya Resident Journalist, February 17, 2010.

Fred Obera, Kenyan freelance journalist, February 18, 2010.

*Keith Somerville lectures in Journalism at Brunel University, London. He is a specialist in radio journalism, ethics, news writing and media coverage of Africa. Prior to joining Brunel he worked for the BBC for 28 years, most of it at the BBC World Service, specialising in the coverage of Africa. He is the author of a number of books and academic articles on conflicts in Africa, including: Southern Africa and the Soviet Union: From Communist International to Commonwealth of Independent States (London: Macmillan, 1993); Foreign Military Intervention in Africa (London, Pinter,1990); Angola: Politics, Economics and Society (London: Pinter, 1986).

55

The author would like to thank Gabrielle Lynch of Leeds University; David Anderson of the University of Oxford‘s Centre for African Studies; Joseph Warungu of the BBC World Service; Matthias Mundi and Chris Greenway of the BBC Monitoring Service; Ida Jooste, Brice Rambaud and the staff at Internews Nairobi; Dennis Itumbi; Martin Gitau; Louise Tunbridge; Fred Obera; and Caesar Handa for their assistance in the research for this paper